Spyros K. Stamatelos1, 2, Arvind P. Pathak2, Aleksander S. Popel2
1Roche Innovation Center, New York, NY, USA, 2Johns Hopkins University, School of Medicine, Baltimore, MD, USA
Objectives: The goal of the project is to develop a computationally intensive model to characterize a whole human tumor xenograft comprised of a population of thousands of vessels. Specifically, the platform cast as blood flow model which processes the output of a graph-based algorithm we developed to delineate blood flow transport in the heterogeneous tumor angiogenic network. This computational model can have wide implications as a simulation tool to design efficacious chemotherapeutic and antiangiogenic strategies [1].
Methods: The xenograft data were obtained from inoculation of human breast cancer cells in the mammary fat pad of mice. The high-resolution images were processed in a graph-based algorithm we developed to reconstruct and repair the morphological discrepancies of the initial image. These imaging data were the input to a blood flow model accounting for pressure drop and blood rheology in all the functional segments of the tumor vasculature. An optimization algorithm was used to ensure mass balance in all the inputs and outputs of the vascular system allowing exchange of blood with an extravascular compartment. Morphological and hemodynamic metrics were calculated for the entire tumor vasculature and various regions of interest. Model development was completed using MATLAB and Java Eclipse programs.
Results: The image-based computational model was successful in processing the whole population of imaging vascular segment data allowing the generation of detailed blood perfusion maps for the whole network and regions of interest. The metrics of both structural (diameter, length) and functional (velocity, hematocrit) characteristics of individual segments were calculated and compared.
Conclusions: Tumor vasculature is extremely heterogeneous and therefore there is a dire need to develop computational tools to characterize it both structurally and functionally [2]. This analysis can facilitate the evaluation of tumor in various stages of progression and provide an assessment tool for targeted therapies.
References:
[1] Stamatelos, S. K., Kim, E., Pathak, A. P. & Popel, A. S. A bioimage informatics based reconstruction of breast tumor microvasculature with computational blood flow predictions. Microvascular research 91, 8-21, doi:10.1016/j.mvr.2013.12.003 (2014).
[2] Kim, E. et al. Multiscale imaging and computational modeling of blood flow in the tumor vasculature. Annals of biomedical engineering 40, 2425-2441, doi:10.1007/s10439-012-0585-5 (2012).
Reference: PAGE 25 (2016) Abstr 5804 [www.page-meeting.org/?abstract=5804]
Poster: Methodology - New Modelling Approaches